The aim of this project is to design and implement an image data compression system using the Classified Vector Quantization (CVQ) technique, in order to get high compression ratios and simple decoding processes. The image to be coded is decomposed into small blocks of size 4 x 4. An edge- oriented classification is used to classify the blocks into classes. For each class an optimal vector quuantizer is developed. The Linde-Buzo-Gray(LBG) algorithm is used to design a subcodebook for each class depending on the VQ technique that used in that class. CVQ can be as a multiple VQs. All basic ideas are scketched and some comments are made on idifications of the basic VQ technique. Also, in this thesis, a complete implementation of CVQ is proposed in four stages. These are building training sets, codebook design, encoder design, and decoder design. Each stage is constructed from several modules. Finally, the system developed in this project, can be used in any of the monochrome image storage/transmission application.